are football coaches overpaid? by randall s....
TRANSCRIPT
1
Are Football Coaches Overpaid?
Evidence from Their Employment Contracts
By Randall S. Thomas* and R. Lawrence Van Horn**
Draft of August 25, 2014
Please Do Not Cite or Quote Without Written Permission of the Authors
*John S. Beasley II Professor of Law and Business, Vanderbilt Law School, Vanderbilt University.
**Associate Professor of Economics and Management, Owen Graduate School of Management, Vanderbilt University.
We would like to thank John Dotson for his outstanding research assistance on this project. We also wish to thank the athletic directors, search firm representatives, agents, coaches, and reporters that we interviewed for their assistance. We have kept their names confidential at their request.
2
Abstract:
The commentators and the media pay particular attention to the compensation of high
profile individuals. Whether these are corporate CEOs, or college football coaches, many critics
question whether their levels of remuneration are appropriate. In contrast, corporate governance
scholarship has asserted that as long as the compensation is tied to shareholder interests, it is the
employment contract and incentives therein which should be the source of scrutiny, not the
absolute level of pay itself. We employ this logic to study the compensation contracts of
Division I FBS college football coaches during the period 2005-2013. Our analysis finds many
commonalities between the structure and incentives of the employment contracts of CEOs and
these football coaches. These contracts’ features are consistent with what economic theory
would predict. As such we find no evidence that the structure of college football coach contracts
is misaligned, or that they are overpaid.
3
Introduction
Corporate Chief Executive Officers (“CEOs”) and Division I Football Bowl Subdivision
(“FBS”) college football coaches share many common characteristics, including that they receive
intense criticism over the level of their compensation. For many years, the mainstream media
has reported on the sensational returns to CEOs. 1 More recently, college football coach
compensation has come under public scrutiny as their compensation has risen.2
Yet economic theory tells us that lucrative compensation is not necessarily a problem if
the structure of the employment contract is aligned with value creation.3 For example, a
significant corporate governance literature has studied the structure of CEO employment
contracts and determinants of compensation, asking if they are aligned with the maximization of
shareholder value.4 Theory suggests that while the dollar value of executive compensation may
be great, as long as it is structurally aligned with value creation for shareholders, the amounts
may be warranted.5 But it is necessary to study the underlying determinants of CEO
compensation embedded in the contract to truly evaluate their effectiveness in creating
shareholder wealth.
Can the same be said about college football coaches? In an effort to shed more light on
this debate, this paper provides a comprehensive evaluation of college football coach 1 Gretchen Morgenson, An Unstoppable Climb in C.E.O. Pay, N.Y. TIMES, June 29, 2013, http://www.nytimes.com/2013/06/30/business/an-‐unstoppable-‐climb-‐in-‐ceo-‐pay.html?pagewanted=all&module=Search&mabReward=relbias%3Ar%2C%7B%221%22%3A%22RI%3A11%22%7D&pagewanted=print . For an academic version of this critique, see LUCIAN BEBCHUK & JESSE FRIED, PAY WITHOUT PERFORMANCE: THE UNFULFILLED PROMISE OF EXECUTIVE COMPENSATION (Cambridge: Harvard University Press 2004). 2 Tamar Lewin, Colleges Increasing Spending on Sports Faster Than on Academics, Report Finds, N.Y. TIMES, April 7, 2014, http://www.nytimes.com/2014/04/07/education/colleges-‐increasing-‐spending-‐on-‐sports-‐faster-‐than-‐on-‐academics-‐report-‐finds.html?module=Search&mabReward=relbias%3Ar%2C%7B%221%22%3A%22RI%3A7%22%7D&_r=0. 3 John E. Core, Wayne R. Guay & Randall S. Thomas, Is U.S. CEO Compensation Inefficient Pay Without Performance?, 103 MICH. L. REV. 1142 (2005). 4 Kevin J. Murphy, Executive Compensation, in 3B HANDBOOK OF LABOR ECONOMICS 2485 (Orley C. Ashenfelter & David Card eds., 1989). 5 Michael C. Jensen & William H. Meckling, Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure, 3 J. FIN. ECON. 305 (1976).
4
employment contracts to uncover the underlying determinants of their observed compensation
and its corresponding alignment with football program value creation. We do so in the same
way that others have evaluated CEO employment contracts in the corporate sector and draw
parallels to that literature where appropriate. We conclude that observed compensation for
college football coaches is the result of their success, based on contracts with performance
incentives and displaying significant levels of employment risk, much the same as with CEOs. In
other words, football coach contracts are very much like CEO contracts in that, while they pay
coaches a lot of money, they are structurally aligned with value creation for football programs so
that the compensation levels may be warranted.
Our article is structured as follows. We start with a brief primer on the economic theory
of employment contracts. Economic theory leads us to make several predictions about the
differences and similarities between college football coaches and CEOs with respect to their
compensation arrangements and contract provisions. First, we predict that college football
coaches will have their pay tied more closely to performance, have shorter employment
contracts, have shorter job tenure, have contracts that focus more on termination events, and
have contracts that contain bigger severance payments for departing coaches. Each of these
predictions arises from the fact that college football coach performance is more easily observed
than CEO performance, and thus we would anticipate that successful coaches would be rewarded
more generously for their successes, while unsuccessful coaches will be more quickly discovered
and terminated.
Next, theory predicts that football coaches will see their compensation rise more rapidly than
CEOs and we would further expect to see total compensation of the coaches to increase at a rate
that exceeds those of CEOs holding all else constant. The explanations for these predictions are
more complex. College football coaches have less firm-specific capital and will therefore be
(generally) more mobile than CEOs. If they are successful, universities will need to offer them
increased pay or risk losing them to other bidders. Since coach success is readily observable,
their contribution to success is clear, and the payoffs to the university for success are quite high,
we would anticipate that coaches would have substantial bargaining power. Further, there are
5
clear hierarchies among universities in the quality of football programs and some top tier schools
are always looking to replace unsuccessful coaches. These factors should in theory lead to
college coaches getting rapid pay increases and a greater percentage of program revenues over
time.
In section II, we compare the recruiting processes for football coaches and CEOs. Based
on interviews with many participants in the hiring process, we make a number of observations.
First, CEOs are frequently internal hires, whereas college football coaches are almost never
internal candidates. Internal candidates have less leverage in contract negotiations than
outsiders,6 which should dampen the level of CEO pay increases relative to that of football
coaches. A second important difference is that the most contentious contracting issue for football
coaches relates to the amount of their severance payments if they are terminated without cause,
or the amount of their contract buyout clause if they leave the university without good reason.
While severance provisions are also important to CEOs, they do not have buyout provisions in
their contracts.
Section III contains detailed comparisons of the different aspects of the two sets of
employment contracts. We compile 947 written employment contracts for head football coaches
at major public universities that play Division I FBS football for the period 2005 to 2013.7 We
then compare the football coach compensation arrangements with a size and revenue matched
sample of public corporations to determine their similarities and differences.
When we examine the two groups’ compensation arrangements and the language of their
contracts, we make several important findings. First, while college football coaches virtually
always have written employment contracts, only 50% of CEOs have such agreements.8 This
6 Stewart J. Schwab & Randall S. Thomas, An Empirical Analysis of CEO Employment Contracts: What Do Top Executives Bargain For?, 63 WASH. & LEE L. REV. 231, 237-‐238 (2006). 7 Generally speaking, public universities must disclose their football coaches’ contracts if asked to provide them under state freedom of information act statutes. For more discussion of our data collection process, see infra Section IIIA. 8 Public corporations have disclosure obligations under the federal securities laws that mandate they publish their CEO’s employment contract in their electronic filings in the EDGAR system. See infra Section IIIA for further discussion of how we obtained these agreements.
6
difference derives from the fact that CEOs hired from another firm are more likely to have
employment contracts than those that are hired internally because “these CEOs tend to face
greater uncertainty about the sustainability of their relationships with their firms than CEOs who
have been promoted from within.”9 Since football coaches are almost always outsiders, this
explains why they are more likely to have written employment contracts. Second, college
football coaches have much shorter job tenure than CEOs do, although their employment
contracts have longer terms on average. Third, football coaches change their contracts
frequently: every two to three years on average, which is far shorter than their average stated
length.
We further observe that college football coach compensation is rising faster than CEO
pay: from 2007 to 2011 CEO pay increased 23%, while over that same time period college
football coach pay increased 44%.10 We find that average fixed pay for our football coaches
more than doubled during our sample period, likely as a result of large increases in television
revenues for many conferences. Part of the reason for the observed differences in compensation
may stem from the fact that football coaches’ success is easily observed, and their contribution to
a school’s win-loss record has a big impact on program value, therefore successful coaches will
be well-rewarded, especially when the overall pie is growing. Conversely, poorly performing
coaches will be quickly terminated, which is the flip side of the tight connection between pay
and performance.
Economic theory suggests that pay for performance would be more prevalent in college
football coach compensation than in CEO pay.11 We find that, while both sets of contracts
typically contain performance pay, in football coach contracts these incentives are relatively
small in comparison to those in CEO contracts. We argue that these features are the result of
coaches having much shorter job tenure than CEOs and therefore engaging in more frequent
contract re-negotiations. Good or poor performance will be promptly reflected in either a new 9 Stuart L. Gillan, Jay C. Hartzell & Robert Parrino, Explicit vs. Implicit Contracts: Evidence from CEO Employment Contracts, 64 J.FINANCE 1629, 1631 (2009). An insider may be better informed about the likelihood that the firm’s board of directors would honor an implicit, non-‐written, employment contract. Id. at 1634. 10 Erik Brady et al., Coaches’ Pay Surges, TENNESSEAN, November 20, 2012, at C1. 11 Jerold Warner, et al., Stock Prices and Top Management Changes, 20 J. Fin. Econ. 461 (1988).
7
improved contract or a prompt termination, making performance based pay less important as an
incentive.
Another important set of differences arises out of how coaches and CEOs are treated
when their performance is poor.12 Coaches are frequently terminated in this situation, and
therefore the conditions surrounding termination are heavily negotiated. With such readily
observable and clean measures of success, this is not surprising. We find that coaches’
employment contracts contain higher severance payments than CEOs, although they are also
more likely to have to make significant buyout payments to their schools if they depart
prematurely to take another head coaching position. We also find that coaches generally have
much more detailed sets of job responsibilities than CEO’s, which may reflect their
accountability to a greater set of stakeholder interests than their corporate counterparts.
Finally, we find some differences surrounding two common and important CEO
employment contract provisions – non-competition provisions and arbitration clauses. Non-
competition clauses are almost entirely absent from football coach contracts, while arbitration
provisions are much less common. Non-competition clauses are designed to protect a company
from the loss of firm specific capital investments in its employees by insuring that the employee
does not unfairly take that investment when they leave the company. Football coaches have little
firm specific human capital investment and little confidential business information so that
universities have little to protect. Moreover, a broadly written non-competition clause would
effectively preclude a coach from participating in the football coaching market, so they resist
them strongly. As a result, universities have generally not included non-competes in coach
employment contracts, except in some cases clauses barring the solicitation of current recruits.
For arbitration, we argue that path dependence has led to it being adopted very slowly in the
coaching area. We finish with some brief conclusions about the implications of our results.
I. The Economics of Labor Markets: Coaches and CEOs
12 Id.
8
While both CEOs and college football coaches are responsible for the performance of their
organizations, economic theory would suggest that based on differences in their contribution to
output, the observability of their success, and the strength of the outside market for their talents,
football coach contracts should differ from those of CEOs.13 In particular, we anticipate that it
will be easier to structure an efficient compensation contract for football coaches than for CEOs.
We begin with the most basic challenges for designing top level executive compensation
contracts. First, an optimal compensation contract should tie executive pay to increases in firm
value.14 There are a variety of metrics that can be used to assess firm performance, including
revenues, sales, profits, and market share, but it is unclear which is the “best” one, i.e. the one
most aligned with shareholder returns and investors’ time preference for returns. Moreover, to
the extent that there is a tradeoff between profits today and long term value, the appropriate
metric is ambiguous. These issues create uncertainty in CEO contracting.
There is much less confusion around metrics to assess value creation by coaches. A coach
creates value for his program when the team wins. This is easily observable and measureable.
Similarly, if we examine the tradeoff between long term value creation and short term value
creation, a football coach who maximizes short term on-field performance normally does so with
little downside risk to the future value of the program. A CEO, on the other hand, may make
inter-temporal substitutions in observed firm performance which are inconsistent with long run
value maximization.
In a related vein, the CEO’s contribution to firm performance is less clear. In a complex
organization like a large public corporation, a multitude of inputs contribute to firm performance,
making it very difficult to untangle the CEO’s marginal contribution to success.15 By
13 Bengt Holmstrom and Paul Milgrom, Multi-‐task Principal Agent Analysis: Incentive Contracts, Asset Ownership and Job Design, 7 J. Law, Econ. & Org. 24 (1991) 14 Hamid Mehran, Executive Compensation Structure, Ownership, and Firm Performance, 38 J. Fin. Econ. 163 (1995); RANDY R. GRANT ET AL., THE ECONOMICS OF INTERCOLLEGIATE SPORTS 224 (World Scientific Publishing Co. 2008). 15 James A. Brickley and Michael S. Weisbach, Accounting Information & internal Performance Evaluation: Evidence from Texas Banks, 17 J. Accounting and Economics 331 (1994).
9
comparison, tying coach compensation to performance is far simpler as coaches’ decision-
making and effort is unambiguously tied to football program performance. Program performance
can be measured without error. The coach’s decisions directly determine whether the football
team either wins or loses and the coach is measured by his win-loss record.16 As such, theory
would suggest that an efficient football coach contract will be more tightly coupled with
performance (greater pay-for-performance sensitivity) than what is observed for CEOs.
Theory predicts that football coach contracts would be shorter in duration as poor
performance can more easily be observed for coaches than for CEOs.17 Losing records lead to
termination, and not all coaches can win, so we would expect short job tenure. In a related vein,
the average coach’s shorter job tenure means that there will be a stronger upfront focus on what
happens to the parties when the coach leaves either voluntarily or involuntarily. From the
school’s perspective, if the coach is highly successful, they want to prevent him from leaving too
easily by increasing switching costs to the university that lures them away. This means that they
will negotiate to require the voluntarily departing coach (and indirectly the school hiring them) to
make substantial contract buyout payments if the coach leaves.
Conversely, an incoming coach is well aware that not all coaches can have winning records
and that the consequence of a losing record is likely to be termination. As a result, the incoming
coach will bargain hard to obtain substantial severance payments if he is terminated without
cause.18 Shorter job tenure should also lead to shorter employment contracts with more frequent
contract amendments granted to successful coaches.
Another noteworthy difference relates to the different levels of firm-specific human
capital for CEOs and coaches. The game of football has a standard set of rules, and recruiting
players requires a fairly standard set of skills, such that winning in one school is likely to be
transferable to another school. In the parlance of labor economics, the coach has general human
capital that is readily moved to another football program. By contrast, given the heterogeneity
around how firms are organized, their production functions, and their products and services, a 16 GRANT, supra note 11, at 228 (“…college coaches are paid, first and foremost, to win.”).
17 Brickley and Weisbach, supra note . 18 An employee can be terminated for cause when they commit an act or omission that is specifically listed as grounds for termination in their employment contract. See infra section IIID for further discussion.
10
CEO will likely require substantial firm-specific knowledge to be a success.19 As a result, a
CEO from one firm has a lower value to another firm because their skills and experience are not
fully transferable. Thus, we would expect to see greater mobility amongst football coaches
across schools than we would see amongst CEOs across corporations. In fact, it is common for a
college football coach to hold many head coach positions at different universities over their
career.
A further point of difference between the two markets relates to the timing of success.
Football has a four month season and the product of the coach’s effort is easily observed. Not so
for the CEO because she is concerned with continuous value creation throughout the year. This
difference affects the timing of the relative performance evaluation in the football coach market.
As we noted above, given the transferability of the coach skill base, recruiting successful coaches
from other schools is common. Each December, at the conclusion of the football season, football
programs review performance and make coach changes prior to National Signing Day in
February. The market for college football coaches clears in this narrow window. As a result,
each program must be prepared to make such switches in short order, having in hand a set of
prospects. Taken jointly, these factors lead to a much more fluid market in which coaches are
likely to have shorter tenure at any given school and are likely to work for many more programs
over their career.
Finally, we expect to see an efficient labor market with a clear hierarchy. Larger football
programs generally have bigger fan bases, greater visibility, and more valuable TV contracts.
All of these features increase the value of the program and the potential payoff for a successful
coach. From a coach’s perspective, this means that running the football program at the
University of Alabama is inherently more valuable than being in charge at the University of
Buffalo. As a result, we observe a natural hierarchy in the labor market where coaches
demonstrate success in lower valued programs and move successively to their highest value use
19 Dawn Harris and Constance Helfat, Specificity of CEO Human Capital and Compensation, 18 Strategic Management Journal 895 (1997).
11
at the larger programs. This illustrates how human capital readily migrates toward its highest
valued use.20
At any given point in time, about half of the top tier schools will have losing records and be
considering replacing their coaches. The job prospects for coaches with losing records are not
good. Given the observability of performance, they have effectively demonstrated their
capability – or lack thereof. The result is that poor performers are not circulated amongst the
ranks of top programs but rather drop out of that coaching pool. This attrition naturally gives
rise to an annual inflow of new talent from lower ranked schools. This means that successful
coaches at lower tier schools are in high demand and have significant bargaining power with
their current employers given the annual need for successful coaches at high value programs.
This concentrates the labor market for talent, imbues successful coaches with market power, and
enables them to capture a greater share of the value they create in their existing programs. If we
couple this observation with the value of the coach’s threat to leave to take his next best
alternative job, and the Division I FBS football programs’ growth in popularity and financial
value over the past decade, we would further expect to see total compensation of the coaches to
increase at a rate that exceeds those of CEOs holding all else constant.
II. Recruiting, Hiring and Negotiating a Contract
To compare college football coaches and CEO’s, we start by examining the process used in
recruiting and hiring them as well as how their contract terms are negotiated. This analysis
provides both a backdrop for understanding the relative bargaining power of the parties involved
and informs our comparison of the terms in their employment contracts.
20 Kevin J. Murphy and Jan Zabojnik, CEO Pay and Appointments: A Market-‐based Explanation for Recent Trends, 94 American Econ. Rev. 192 (2004).
12
A. College Football Coaches21
Historically, on average, every three years a Division I FBS university will change its head
football coach.22 This event is costly as Tennessee’s recent divorce from its head football coach
Derek Dooley illustrates: after three years and a 15-21 win loss record, the university terminated
his contract but is obligated to pay Dooley $5 million severance pay as well as potentially paying
several of his assistant coaches another $3.71 million.23 Universities therefore have powerful
incentives to identify successful coaches when they have to find a new one. One of the best
predictors of success in a new job is prior success in a similar position so that there is strong
competition to attract winning coaches.
1. Search Firms and Generating a List of Names
The starting point for the football coach hiring process is an announcement that the current
head coach at Major State University (“MSU”) has been fired, announced his resignation to take
another head coaching job, or decided to retire. An experienced athletic director (“AD”) has
either anticipated this announcement for some time (in the case of retirement or termination by
the school), or is aware that the head coach has compiled a winning record for MSU over the
past one or two seasons and therefore suspects that the coach will be contacted about job
offerings at more prestigious (or richer) programs.24 As a result, the AD has at all times in his
21 This section is based on confidential discussions with search firm representatives, agents, coaches, reporters and athletic directors. We use the pronoun “he” throughout because almost all of the participants in this process are male. 22 Martin J. Greenberg, College Coaching Contracts Revisited: A Practical Perspective, 12 MARQ. SPORTS L. REV. 127, 129 (2001). 23 David Climer, Tennessee Racks Up Buyouts, Not Wins, TENNESSEAN, November 21, 2012, at C1. 24 Head coaches typically ask permission from the athletic director to speak with other schools before interviewing with them for other positions. However, some coach employment contracts go further and obligate the coach to ask permission from his employer before talking to other schools. John Taylor, Alvarez: contact disclosure will be
13
pocket a list of three to six names of potential replacement coaches that have for one reason or
another caught his eye. This list will contain candidates who have either expressed an interest,
usually through an agent, who are doing well at a less-highly regarded football school, or who
are coveted, well-known coaches at established programs. It is rare for this list to include any
internal candidates.25 The AD knows that the process will move quickly once it is apparent that
the current head football coach is leaving. However, the AD’s have agreed informally that they
will not contact coaches at other schools until their season is finished, or their bowl game is
completed, and they must name a new coach before potential recruits announce their enrollment
decisions in February. If it delays, the school may lose an entire year’s recruiting class. This
means that there is intense pressure to fill the vacancy quickly.
At this very early stage in the process, MSU is likely to engage a search firm to assist with
the hiring process. There are a small set of search firms that handle coaching searches. These
firms act as intermediaries between the schools and the potential candidates or their agents. The
search firms are continually updating their information about what coaches are available, which
coaches might be interested in which jobs, and which schools might in turn be interested in them.
In addition to gaining access to the search firm’s valuable insights and contacts, MSU employs
the search firm to ensure secrecy and to honor the informal agreement among fellow AD’s that
they do not solicit one another’s coaches without prior permission. It is important that MSU
does not directly contact the candidates, so that it can preserve its ability to truthfully deny that it
has contacted any of these prospective coaches.
part of next UW coach’s contract, NBCSPORTS.COM, December 10, 2012, http://collegefootballtalk.nbcsports.com/2012/12/10/alvarez-‐contact-‐disclosure-‐will-‐be-‐part-‐of-‐next-‐uw-‐coachs-‐contract/. 25 One reason why internal candidates are rare is that assistant coaches’ success is very visible to other schools, and the hiring interval is very short for head coaches, so that top quality “heir apparents” often get snapped up by other schools. For example, James Franklin was an assistant coach at the University of Maryland where he had been anointed the likely successor for the head coaching position. See Jeff Barker, Friedgen’s firing was weeks in the making, BALT. SUN, Dec. 21, 2010, http://articles.baltimoresun.com/2010-‐12-‐21/sports/bs-‐sp-‐terps-‐fridge-‐1221-‐20101220_1_lame-‐duck-‐atlantic-‐coast-‐conference-‐coach-‐football-‐coach. However, when Vanderbilt University needed to hire a new head football coach, they succeeded in hiring him away. Id.
14
Once they are contacted, the search firm will ask the AD to fill out a detailed questionnaire
about the prospective coach’s anticipated financial package and job responsibilities. This
document is compiled over time by the search firm based on the questions that they have
received from coaches in prior searches. Using this detailed information, the search firm will
take the AD’s short list of prospects and add to that list other candidates they have identified
until they have a complete list of possible candidates.
The search firm will then begin their conversations with prospective coaching candidates.
These talks are of a preliminary nature until the potential candidate’s team has finished its
regular season, or its bowl game. The discussions provide potential candidates with information
about the range of their likely compensation as well as the names of the players that the school
has recruited, the size of the pool of assistant coach compensation, the likely length of assistant
coach contracts, and several other key preliminary issues. The search collects information about
the potential candidates’ existing employment contracts, especially the size of any buyout
payment that MSU (or the coach) is likely to have to pay to the coach’s home school if he should
leave. Some candidates may be dropped at this point if the school cannot meet their
requirements.
2. The Coach’s Agent’s Role
At this stage, a candidate’s agent enters the process.26 While not all coaches have agents, for
those that do, the agent will provide a variety of services. Oftentimes an agent will have ongoing
communications with the search firms to give them the names of potential candidates and to get a
sense of the overall market, even when there is not a particular search underway. Once a job
becomes open, the agent will often become more aggressive in promoting his clients as potential
applicants, sometimes directly contacting the AD to suggest that MSU consider a particular
26 Agents are compensated by the coach, not by the school. They have become more commonly used over
time.
15
client represented by the agent. The agent may also act as the intermediary in early discussions
with the search firm.27 The agent will negotiate, on the coach’s behalf, issues such as
compensation, staff levels/pay and football program resources.
Once the search firm moves beyond its initial assessment of the potential candidate’s interest,
and before it engages in any substantive recruiting discussions, the potential candidate will
normally request permission from his AD to talk with MSU. Generally speaking, universities
will permit their current coaches to talk to other schools at the appropriate time after their season
is over.
The search firm will, in consultation with MSU, now have prepared a memorandum of
understanding (“MOU”) for the candidates. The MOU lists their salary, the length of their
contract, their perquisites, incentive pay, guaranteed payments from television, radio and shoe
contracts, their termination benefits and their buyout clause. These terms are frequently
“benchmarked,” that is, compared with those offered to coaches at comparable schools (such as
those in the same conference).
By the end of the initial search process, the potential candidate will have clarity around what
compensation package the school is offering, what staff they can hire, the school’s resource
commitment to the football program, its’ current recruits, and whether there is an ideological fit
between the coach and the school. MSU, on the other hand, will have a reasonable idea about
the candidate’s level of interest in the job and the potential fit between the school and the
candidate.
3. Initial Interviews
27 Agents will counsel their clients about the potential desirability of participating in MSU’s search, and if the
pool of potential candidates is too large may tell them to avoid it so as not to hurt their future marketability.
16
If the initial discussions between the search firm, the agent and/or the potential candidate,
indicate that there is a likely fit for both sides, then the school will schedule an initial interview
with the candidate. Typically, the school will hold this first round of interviews off campus with
six to ten candidates with an eye toward inviting a few of them back to campus for a second
round of interviews. Agents do not participate in these interviews.
At the initial interview, a small group of the school’s representatives will ask the candidate
about his coaching philosophy and his abilities as a recruiter, as well as try to further assess how
interested the candidate is in the position. The candidate will sell himself to MSU in hopes of
getting invited back to campus for a second interview. While many candidates are genuinely
interested in the position at this point, some coaches may use these discussions as a negotiation
ploy with their home school for a raise and a new or amended employment contract.
Once this first round of discussions is completed, the MSU representatives, including the
AD, his staff and often a few prominent alumni, will engage in internal discussions regarding the
candidates. They will also contact outside sources of information to collect as much information
as possible about the candidates, including background searches to uncover any potential
problems with the candidate. Generally about 50% of the candidates will drop out of the process
at this stage.
4. On Campus Interviews
Based on the information available to MSU by the end of this process, the school will invite
three or four candidates to on campus interviews. Generally speaking, all of these candidates the
school would be happy to hire as its new head coach. Furthermore, unlike the first round of
interviews where the candidates were selling themselves, in the second round of interviews, the
school is selling itself to the candidates. No agents are present at these meetings.
17
On campus interviews are much more involved than the initial interviews. The candidates
will meet with MSU’s President, some trustees and important administrators, in addition to the
AD and his team. If married, they are likely to be asked to bring their wife along to see the
university’s locale. Based on these meetings, and the input from key alumni, administrators, and
the President, the AD will decide on MSU’s order of preference for the candidates. Once the
order has been established, the school will continue to negotiate in secret with several candidates
simultaneously until they are confident that their preferred candidate will accept the job. As we
predicted in section I, the most contentious issue to negotiate is what the coach will receive in
compensation if they are terminated by the school and what they will pay the school if they
choose to leave the job for another position.
By the end of these negotiations, each serious candidate will know what to expect from MSU
if they are offered the position and, if they are the frontrunner, are likely to have been given a
letter of intent. Ultimately, MSU will not offer the job to any candidate without assurance the
candidate will accept the position, so as to avoid the embarrassment of having a candidate turn
down the job and thereby revealing to the remaining candidates that they were not MSU’s first
choice.
5. Accepting the Job and Afterwards
Once the job is accepted by the candidate, the agent will complete the negotiations with the
school. The agent’s job is to translate the MOU into a final long form contract that insures that
his client is protected against unfair termination, or from being asked to engage in overly
demanding sets of activities. The main focus of these contract negotiations will be on the
definition of termination for cause and the head coach’s enumerated duties in their contract, yet
other details such as the form of incentive pay will be determined. Some of these issues can be
resolved based on the school’s contracts with prior coaches, or by reference to clauses from the
contracts of coaches at other schools. This final set of negotiations typically take anywhere from
18
two to six months depending on the urgency of the issues and the parties’ positions, but in most
cases, the coach has substantial negotiating leverage because the school wants to insure that the
coach has a successful start to their tenure.
If the coach is successful at MSU, it is likely that he, or his agent, will ask the university to
renegotiate his contract after a couple of years. These contract renegotiations lead to contract
amendments, which are usually simple one page documents that enumerate any changes to the
contract’s length, the coach’s base salary or other guaranteed income, any additional bonuses,
and any changes to the buyout amount. These changes are again benchmarked against the
amounts paid to other coaches in the conference. Given the coach’s success, it is likely that he
will have substantial negotiating leverage in this negotiation and can extract significant
concessions from the university.
B. Negotiating CEO Employment Contracts28
The CEO recruitment process has both similarities and differences from the college
football coach recruitment process. When an existing corporate CEO retires, dies, or is fired, the
firm must find a replacement. Unlike the coach process, the day-to-day management of the
search will generally be run by an ad hoc committee of the board of directors, subject to overall
review by the full board. Most firms will look first internally to see if they can find a
replacement. For an internal search, the directors will keep the process in-house and not engage
an executive search firm, although they may use a compensation consultant to advise them on
pay structure and levels.
28 This section is derived from earlier work by one of the authors. Schwab & Thomas, supra note 6, at 236-‐240. As noted there, it is based on discussions with legal practitioners that were actively involved in negotiating executive employment agreements, sometimes on behalf of companies, and sometimes on the side of the executive. As in the earlier piece, we have continued to conceal their identities to preclude any possible adverse use of this information.
19
If outside candidates are being considered, the search committee will usually retain an
executive search firm. Together the directors and the search firm will generate a group of
potential candidates for the position.29 After reviewing more information, the search committee
will narrow the list down to three to five finalists. These candidates will be interviewed, often
secretly so that their current employers will not be aware of their potential interest.
The search committee will generally offer the position to the top choice. They will provide
the candidate with a term sheet setting forth the proposed salary, target bonus, equity
participation in the company, severance package, change-in-control protections, benefits and
standard perquisites. Compensation levels will frequently be benchmarked against comparable
firms. The term sheet will also cover such items as relocation expense payments, the term of the
years for the contract, the renewal provisions for the contract, the duties associated with the
position, and the physical location for the executive. Many times the term sheet is the only
written documentation of the employment arrangements, although about half the time the parties
enter a written employment contract.30
After completing negotiations between the principals over the term sheet, the two parties’
attorneys will focus on the final terms of the employment contract. The company normally will
provide an initial draft of the contract, often based on their prior CEO’s contract. The company
has an advantage because the final contracts will become public information and therefore the
company wants to avoid establishing unfavorable future precedent for later negotiations. Only in
rare situations where a candidate has extraordinary negotiating power would the firm allow them
to offer an initial draft. The company may use inside counsel, or in some cases outside lawyers,
but the executive will always have their own lawyer to handle this part of the transaction. Some
lawyers specialize in representing top executives in their job negotiations and have well-
established reputations for getting their clients the best deals.
C. Principal Differences in the Hiring Processes and Their Effects 29 RAKESH KHURANA, SEARCHING FOR A CORPORATE SAVIOR: THE IRRATIONAL QUEST FOR CHARISMATIC CEOS 28 (Princeton University Press 2002). 30 Gillan, et al., supra note 9, at 1631.
20
One very important difference between the recruiting processes for coaches and CEOs is that
head football coaches at large universities are rarely recruited internally. By comparison, it is
commonly the case that a public company will first look internally to find its new CEO and only
where the talent pipeline is empty will the board turn to outsiders.31 Compared with inside
officers that have been promoted, executives that have been brought in from outside of the
company are typically paid a premium. This occurs because the firm has no viable alternative
internal candidates, which gives the outsider significant negotiating leverage in the job
negotiations.
However, once the candidate has established themselves on the job, these differences in
negotiating power likely diminish. Incumbent CEOs generally have stronger negotiating
positions than incoming CEOs for many reasons including, most obviously, they have already
experienced some degree of success in running the company and they have an ongoing
relationship with the incumbent board of directors. Within a few years, they may be able to use
their new leverage to persuade the board to amend their initial employment contract, or even
offer them a new contract.
In contrast, there are few internal candidates in the football coach market. Coaches have little
firm specific capital, as their football skills are readily transferable, so they can easily move to
another school. Moreover, successful head football coaches have substantial negotiating power.
If their teams win on the field, other schools are likely to invite them to apply for their
openings.32 If their home school is committed to retaining them it will be obligated to offer them
better terms in an amended or new contract. We would therefore expect that the compensation
levels of successful coaches would increase rapidly, as would that of a successful CEO.
A second important difference is the role played by agents in the process. Many coaches or
potential coaches have agents and these agents are often specialized in the football coach market. 31 CEO Inside Promotions Versus Outside Hires, (The Conference Board), May 12, 2014 (finding that in 2013 S&P 500 firms hired outsiders to fill their CEO position only 23.8% of the time). 32 GRANT, supra note 10, at 234 (“Universities, especially those attempting to upgrade their athletics program, often compete against one another to hire the “messiah coach,” the person who will transform a losing program, or a relatively unknown program, into a high profile program where winning is the rule, not the exception.”).
21
A successful agent may significantly enhance their clients’ bargaining power if they are truly
knowledgeable about the level and type of compensation available to candidates in the market.
While some AD’s are experienced repeat players in the game, an inexperienced AD may be at a
disadvantage in negotiating with a big name agent.
Outside CEOs will also sometimes engage lawyers that specialize in representing top
executives. If a search committee is not careful, they may find themselves at a disadvantage in
the contract negotiations if they choose to use inside counsel, or non-specialist outside counsel,
to represent the company.33 However, this is much less likely to occur when the board selects an
inside candidate. As noted above, inside candidates are frequently selected at public companies,
so that candidates’ agents/lawyers are less likely to impact the overall compensation packages,
on average, at firms than for football coaches.
There are a number of other differences in the recruiting processes. For example, coaches
generally seek permission from their schools before they officially talk with other universities
about job openings, while CEOs do so in secret. However, they do not appear to have as
significant an effect on the outcome of the negotiation process as the ones described above.
III. Football Coach Contracts in Comparison with CEO Contracts
While university presidents are nominally the CEO of the university, there are many
commentators,34 including some presidents, 35 who believe that the football coach retains the role
33 GRAEF S. CRYSTAL, IN SEARCH OF EXCESS: THE OVERCOMPENSATION OF AMERICAN EXECUTIVES (Norton 1991). 34 Joe Posnanski, Alabama Coach Nick Saban Unrelenting In Pursuit Of Perfection, NBCSPORTS.COM, November 8, 2013,http://www.nbcsports.com/joe-‐posnanski/alabama-‐coach-‐nick-‐saban-‐unrelenting-‐pursuit-‐perfection (football coaches have “a CEO’s job with ultimate responsibility in marketing and public relations and media relations and community outreach and education and recruiting and game planning and fundraising and budgeting and countless other details.”); Andrew McKagan, Why Ed Orgeron is the Perfect Fit at USC, NEONTOMMY.COM, November 21, 2013, http://www.neontommy.com/news/2013/11/usc-‐orgeron-‐perfect-‐fit?page=2 (quoting former USC head coach Ed Orgeron as saying, “I’m the CEO of the program”); Martin J. Greenberg & Thom Park, CEOs in Headphones, THE SPORT JOURNAL, February 12, 2014, http://thesportjournal.org/article/ceos-‐in-‐headphones (“Today’s major college coaches are CEOs in Headphones.”).
22
as the most powerful decision maker. Football coaches have many of the same job characteristics
as CEOs of public companies – they run large organizations with many employees that generate
hundreds of millions of dollars in revenues and are paid millions of dollars a year,36 frequently
far more than their ostensible bosses, college presidents.37
But are they overpaid? To answer this question, we turn next to a close comparative analysis
of CEO and football coaches’ compensation arrangements and contract terms. If they exhibit
similar characteristics, then we can draw on prior work concerning CEO pay to conclude that
football coach compensation and contractual arrangements have comparable virtues.
A. Description of the sample
Our initial step in obtaining the employment contracts for the college football coaches
represented in our sample was to compile a list of the coaches at each school during the relevant
time period. Once we had created this list, we began looking for these coaches’ contracts in the
online contract databases put together by USA Today.38 Additional contracts were obtained from
various websites through Internet searches using variations of the coach’s name and the words
“employment contract” or “employment agreement.” We then turned to news stories and Internet
articles about contract extensions, amendments, and renegotiations with the various coaches to 35 Doug Lesmerises, Gordon Gee: ‘I think everyone won’ – Closing the book on the Ohio State tattoo scandal, CLEVELAND.COM, June 3, 2014, http://www.cleveland.com/osu/index.ssf/2014/06/gordon_gee_i_think_everyone_wo.html (“Gordon Gee’s involvement with Ohio State’s NCAA issues always will be remembered with eight words: “I’m just hopeful the coach doesn’t dismiss me.”). 36 Todd Brown et al., Performance Measurement & Matching: The Market for Football Coaches, 46 Q.J. BUS. & ECON. 21, 24 (2007) (Although a college football coach serves “under a university president and the athletic director, the responsibility for the conduct, success, or failure [of the team] rests with the head football coach.”) 37 Martin J. Greenberg & Thom Park, CEOs in Headphones, THE SPORT JOURNAL, February 12, 2014, http://thesportjournal.org/article/ceos-‐in-‐headphones 38 The links for the USA Today databases are as follows: 1) 2006 and 2007: http://usatoday30.usatoday.com/sports/graphics/coaches_contracts07/flash.htm; 2) 2009: http://usatoday30.usatoday.com/sports/college/football/2009-‐coaches-‐contracts-‐database.htm; 3) 2010: http://usatoday30.usatoday.com/sports/college/football/2010-‐coaches-‐contracts-‐database.htm
23
ensure that our sample was as complete as possible. If gaps were discovered in the data, we
contacted the university employing the coach and requested the missing contracts. This was done
through a written request pursuant to the freedom of information statute in the relevant states.
We were able to obtain complete sets of employment contracts from 2005 to 2013 for the
coaches at 106 of the 127 Division I college football programs. The majority of the schools for
which we were unable to obtain employment contracts are private universities and, thus, are not
subject to any freedom of information statute. The remaining public universities failed to respond
for various reasons as described below.39
There are four different types of contracts included in our sample. The first group is labelled
employment contracts. These are the initial contracts that the coach and the university entered
into when the coach took the position for the first time. There are 495 of these in our sample and
almost every coach in our sample enters into one of these contracts.40 The second largest
category of contracts is amendments to employment contracts. These are subsequent agreements
that the coach and the university execute after the coach has already been working for the
university for a period of time. Typically, these amendments will update the coach’s
compensation arrangements with the university in the later years of their initial contract. There
can be multiple contract amendments over time to a coach’s initial contract. The third type of
agreement is an amended and restated agreement. These will normally occur once a coach has
entered into several contract amendments to his initial contract and the parties decide that it
would be appropriate to incorporate all of those changes, plus potentially additional changes, into
a completely new contract. Finally, in a small set of cases, the coach and the school will enter
into some other form of agreement that contains some employment terms, as described more 39 The schools with Division I college football programs located in Pennsylvania, Penn State, Temple, and the University of Pittsburgh, are “state-‐related” institutions and fall outside the purview of Pennsylvania’s Freedom of Information Law. See Rob Arcamona, What documents can a reporter obtain from Penn State officials?, STUDENT PRESS LAW CENTER, Nov. 15, 2011, http://www.splc.org/wordpress/?p=2864. Some schools, such as the University of Mississippi and Mississippi State, have a very basic contractual arrangement with the coach. The real contract, containing the bulk of compensation and incentive pay, is entered with a private foundation that is not subject to the state Freedom of Information Law. A few schools, such as the Air Force Academy, simply failed to send the requested employment contracts, despite our repeated requests.
40 As noted below, a handful of coaches entered into some other form of contractual agreement.
24
fully in the note below.41 In total, we have 947 contracts of these four types in our sample. Table
1 below provides some summary information about these contracts.
Table 1: Description of the Football Coaches’ Contracts Sample
Number of Football Coach Contracts by Contract Type
Number Percentage of Sample Contracts
Employment Contract 495 52% Amendment to Employment Contract
397 42%
Amended and Restated Employment Contract
42 4.50%
Other Agreement Including Employment Terms
13 1.50%
Totals 947 100%
Compared with CEOs the most striking difference is that many CEOs do not have written
employment agreements. One recent study found that “less than half of the firms in the S&P 500
had a comprehensive written (or explicit) employment agreement with their CEOs.”42 However,
other research has shown that “the vast majority of the other CEOs [without employment
contracts] had at least some other contractual agreement relating directly to their employment
with their company.”43 Examples of these other agreements are: severance agreements including
change-in-control agreements, consulting agreements, and non-competition agreements, as well 41 Some examples of "Other Agreement including employment terms" are: a Memorandum of Understanding; An Offer Letter; or a Letter of Appointment. A Memorandum of Understanding or Offer Letter is frequently signed before the formal employment contract. They would govern the employment relationship while the formal contract is being negotiated. For this reason, we consider them as equivalent to temporary employment contracts. Letters of Appointment are rare but appear to be abbreviated versions of employment contracts.
42 Gillan, et. al., supra note 9, at 1629. 43 Schwab & Thomas, supra note 6, at 241.
25
as many compensation-related agreements including deferred compensation agreements, stock
option plans, and long term incentive plans. But these agreements generally provide much less
protection for these CEOs than full employment contracts.
Why is there a difference? An earlier empirical study of why CEOs have written
employment contract found that CEOs hired from another firm were more likely to have
employment contracts than those that were hired internally.44 The authors hypothesized that
“these CEOs tend to face greater uncertainty about the sustainability of their relationships with
their firms than CEOs who have been promoted from within.”45 This would also seem to be true
for football coaches who are hired from outside the firm and whose success (and failure) is more
readily observed than with CEOs. Moreover, CEOs change jobs very infrequently relative to
college football coaches. Both universities and coaches have strong interests in careful planning
for the coaches’ more frequent job changes. This may be coupled with universities need to have
well-specified rights to terminate coaches for NCAA rule violations, as discussed below in
section IIID. Both of these factors would lead to college football coaches being more likely to
have written contracts.46
During the eight year time period, we find the median number of contracts by school is
three, while the mean number of contracts by school is four. This suggests that football coach
contracts change approximately every 2-3 years. Based on our data, college football coaches
have an average job tenure of 4.6 years with the median length equal to three years.
Undoubtedly the main reason for this short tenure is the strong pressure that college football
coaches face to have winning records and, at some schools, to be invited to post-season bowl
games. This pressure can lead to a successful coach leaving to go to a better job, or alternatively,
to an unsuccessful coach being fired quickly.
44 Gillan, et. al., supra note 9, at 1631. 45 Id. An insider may be better informed about the likelihood that the firm’s board of directors would honor an implicit, non-‐written, employment contract. Id. at 1634. 46 Another factor may be that coaches have relatively low levels of firm specific investment in their employers and so it is less expensive for the university to replace their skills with another hire. An employee that is in a position that requires more general skills will prefer a written employment contract. Id. at 1634.
26
In terms of job tenure, CEO tenure overall47 has declined in recent years: in 2000 CEOs had
average job tenure of 9.9 years, but by 2012 this number had declined to 8.1 years.48 A variety of
reasons have been suggested for this decline, including the increasingly competitive global
marketplace, the presence of alternative employment options at private equity firms, and
increased director independence/shareholder scrutiny leading to greater executive
accountability.49 But the striking fact is that, even with this decline, CEOs have far greater
average job tenure than college football coaches, most likely because it is much easier to observe
the win-loss record of a coach than the marginal effect of a CEO on the value of a public
company.
B. CEO and Football Coach Compensation
It is not easy to draw direct comparisons between coach and CEO compensation due to the
differences in organization size and value. To create a useful comparison, we matched football
teams with comparable firms using firm size and revenue. This enabled us to create a matched
cohort of football coaches and CEOs.
For football programs, we draw on the work of others. Professor Brewer computed
valuations of college football programs using a variety of factors related to each program. 50
Brewer’s valuations consider “long-term on-field performance, stadium size, state-by-state
growth rates, cash flow, revenues and other factors.”51 We also examined valuations done by
47 The average CEO of an S&P 500 company is more likely to be terminated after poor job performance or because they have reached retirement age. Company Performance and Age as Determinants of CEO Succession (The Conference Board), June 2014, at 1. However, the likelihood of a CEO losing his or her job because of poor performance has not increased over the past forty years. Id. 48 Departing CEO Tenure (2000-‐2013) (The Conference Board), June 2014, at 1. 49 Id. 50 Anthony Schoettle, Notre Dame No.2 in ranking of football riches, IU tops Purdue, IBJ.COM, Jan. 9, 2014, http://www.ibj.com/the-‐score-‐2014-‐01-‐09-‐notre-‐dame-‐no-‐2-‐in-‐ranking-‐of-‐football-‐riches-‐iu-‐tops-‐purdue/PARAMS/post/45498 (reporting results of research by Professor Ryan Brewer of Indiana University). 51 Id. For example, he found that the highest valued program was the University of Texas at $875 million in 2013.
27
Forbes magazine based on Department of Education data on team revenues and expenses.52
Using these two methods, we determined that we should limit the sample of comparable firms to
those firms with a market cap of less than $1 billion in the Russell 2000.
However, we found that there was significant heterogeneity amongst this subset of the
Russell 2000. In some companies, their market value was tied to revenues, while in others
valuations reflected the market’s expectations of future growth. In the case of football programs,
value is tied most closely to revenues and those streams are fairly predictable over time.
Therefore, as a second screen for what constitutes a comparable firm, we limited our sample to
those firms for whom firm revenues were such that their market capitalization was between 1.5
and 10 times firm revenue. We derived this criteria from our comparison of the relationship
between football program revenue (reported to the Department of Education) and implied market
value of each program by Professor Brewer.
1. Football Coach Fixed Compensation Data
We begin with our data on college football coach contracts and extract the compensation
elements. We define these variables as follows. Salary is the amount of base salary specified in
the employment agreement. If the base salary changes over time, we use the initial base salary.
Given the relatively high frequency with which contracts are revised, this seems a reasonable
decision.
Service compensation is fixed pay that a coach receives from radio, television, internet,
speeches, alumni club appearances, and similar engagements. We report the amount of service
compensation that is listed in the employment agreement and use the initial amount if it changes
52 Chris Smith, College Football’s Most Valuable Teams 2013: Texas Longhorns Can’t Be Stopped, FORBES.COM, Dec. 18, 2013, http://www.forbes.com/sites/chrissmith/2013/12/18/college-‐footballs-‐most-‐valuable-‐teams-‐2013-‐texas-‐longhorns-‐cant-‐be-‐stopped/.
28
over time. In some cases, we are unable to disaggregate service compensation and base salary,
and in these situations, we report the total as base salary. This may lead us to underreport service
compensation in some cases, but we note that it does not affect the total compensation values.
Camp measures fixed pay that a coach receives for participating in a summer football camp.
This usually occurs in one of two ways: either the coach will be paid by the university for his
involvement in a camp conducted by the university; or the coach will conduct the camp, receive
the revenue, and pay the expenses (and sometimes pay the university for the use of university
facilities).
The last major category of fixed compensation is Endorsement. It is common for the
university to pay the coach for his duties pursuant to equipment, shoe, apparel and other
endorsement contracts entered by the university. As with the other fixed compensation
variables, we use the initial dollar value stated in the employment agreement even if it changes
over time.
Table 2A reports Division I FBS football coach fixed compensation, that is, compensation
that would be paid to the coach irrespective of his performance. Given the rapid rate of increase
in football coach compensation over time, we calculated these values for three different years:
2005 (the beginning of our sample period), 2009 (the midpoint of our sample period) and 2013
(the final year of our sample period). During this time period, average fixed football coach
compensation more than doubled.
Table 2A: Division I FBS Coach Compensation: Average Fixed Pay
Contract Year 2005 Contract Year 2009 Contract Year 2013 Salary 252,974 405,631 547,167 Service 396,486 646,741 900,277 Camp 1,964 9,865 4,311 Endorsement 73,750 76,047 119,067 Total Fixed 725,174 1,138,285 1,570,823
29
Compensation Salary % Total 34.88% 35.63% 34.83% Service % Total 54.67% 56.81% 57.31%
A brief examination of the data in Table 2A shows the importance of service income as a
component of coach pay. Over the 2005-2013 time period, service income rose from slightly
less than $400,000 a year to more than $900,000 a year. This is by far the largest component of
college football coach compensation. It also constitutes an increasing percentage of total
compensation over time, rising from 54.67% of total compensation in 2005 to 57.31% of total
compensation in 2013.
Television revenue is one of the key drivers in the recent escalation of service income and
therefore more generally of college football coach compensation. A good example of this is the
compensation paid to Mike Leach by Washington State University. Washington State is a
member of the Pac-12 Conference, which signed a new television deal in 2011. This new deal
equated, for Washington State, to a near quadrupling of television revenue.53
Prior to 2011, Washington State had paid its previous football coach, Paul Wulff, roughly
$600,000 per year. Leach was hired before to the 2012 season (but after the TV contract was
signed) and was paid roughly $2.25 million per year. The Athletic Director of Washington State,
Bill Moos, attributed the University’s ability to pay this drastic increase in compensation directly
to the new television deal, “Without the TV revenues, Washington State simply wouldn’t have
been able to hire a coach of Leach’s caliber.”54
UCLA is another example of the importance of television revenue as a contributor to the
rise of college football coach compensation. UCLA is a member of the Pac-12 Conference and
beneficiary of the new TV deal signed by the Pac-12 Conference in 2011. UCLA Athletic
53 Greg Bishop, Television Revenue Fuels a Construction Boom in the Pac-‐12, N.Y. TIMES, November 29, 2013, http://www.nytimes.com/2013/11/30/sports/ncaafootball/television-‐revenue-‐fuels-‐a-‐construction-‐boom-‐in-‐the-‐pac-‐12.html?_r=0. 54 Lewis Kamb, Big-‐time coaches score big-‐time perks, SEATTLE TIMES, November 23, 2013, http://seattletimes.com/html/localnews/2022321329_coachespayxml.html.
30
Director Dan Guerrero commented on the importance to coach compensation of this increased
television revenue.
“You go back to 2003, when I made my first [football] hire, the economic situation was completely different,” Guerrero said. “That was a constraint, a barrier, to being competitive. We couldn’t snap our fingers and come up with dollars.”. . . . “It’s a big game-changer across the conference,” Guerrero said about the television revenue. “The coaches who have been hired, the practice facilities, for basketball and football, are due in large part to that windfall.”55
These television contracts disproportionately favor the larger, better known conferences. In
particular, the conference that wins the BCS will gain more public attention and therefore receive
a big increase in their revenues when they renegotiate the conference’s omnibus television
contract in the future. Such an increase will benefit all of the conference schools because of
revenue sharing arrangements.
We also present 2013 data on football coach compensation by program size in order to
illustrate some of the differences between how coaches are paid at different size schools. Table
2 B provides this data broken out by quartiles from smallest schools to largest ones.
Table 2B: Head Football Coach Fixed Compensation ($)
Type of Fixed Pay
Smallest Quartile Second Quartile Third Quartile Largest Quartile
Salary 281,303 517,722 481,714 903,409 Service 125,321 481,704 1,015,983 1,498,872 Endorsement 6,206 61,252 48,500 252,704 Total Fixed Compensation
412,832 1,074,679 1,552,447 2,664,955
55 Chris Foster, Money is no longer an object for UCLA football, L. A. TIMES, December 7, 2013, http://articles.latimes.com/2013/dec/07/sports/la-‐sp-‐1208-‐ucla-‐football-‐20131208
31
Schools in the largest quartile of the distribution pay their coaches more than six times as
much as those in the lowest quartile. The difference is particularly stark for service income,
whose main component at the bigger schools is television revenue. It is no wonder that coaches
who are successful at smaller schools move to the larger universities when they get the
opportunity.
2. Football Coach On-Field Performance Based Pay
We next examine the pay for performance components of coach pay, beginning with those
payments that are triggered by on-field performance by the team. There are a series of on-field
performance measures that can be included in these contracts depending on appearances and
games won. Here we report the main categories: play in conference championship; win
conference championship; appear in bowl game; appear in BCS bowl game; win BCS bowl
game; appear in national championship; and win national championship.
Table 2C reports the percentage of annual salary amounts (as stated in the year of contract
initiation) of these bonuses in those contracts that contain them as well as in brackets the
percentage of contracts that have these particular provisions. Obviously some of these events are
more likely to occur than others—for example, only one team can win the national championship
game. Furthermore, not all of these provisions appear in all of the football coaches’ contracts.
So for most coaches the expected value of these provisions is far less than the face amount listed
in Table 2C. Further to win the national championship, a coach would normally have to win
their conference championship and appear in a BCS bowl game. In this situation, the effective
payout becomes much greater.
Table 2C: Average College Coaches’ On-Field Incentive Contract Provisions as a Percentage of Base Salary (Percentage of Contracts Including Term)
32
Contract Year 2005 Contract Year 2009 Contract Year 2013 Play in Conference Championship
18% ( 36% ) 18% ( 37% ) 14% ( 67% )
Win Conference Championship
25% (52% ) 28% (75% ) 23% ( 83% )
Appear in Bowl Game
13% ( 82% ) 16% (88% ) 18% ( 90% )
Appear in BCS Bowl Game
37% (59% ) 36% ( 71% ) 39% (83% )
Win BCS Bowl Game
77% ( 4% ) 76% ( 10% ) 52% ( 27% )
Appear in National Championship
53% (55% ) 42% (74% ) 50% (80% )
Win National Championship
95% ( 45% ) 87% ( 48% ) 94% (63% )
There is a clear trend toward greater use of these provisions over time: every category of
on-field performance based pay clause appeared more frequently in 2013 than it did in 2005.
However, their relative size compared to base salary in CEO contracts has remained virtually
constant over the same time period.
Finally, it is worth noting the interdependence of these provisions. In any one conference,
there can only be two finalists in the conference championship and only one of them can win the
game. While it is true there are a large number of bowl games, it is also true that there are a very
limited number of BCS bowls and most of the participants in them are likely to be conference
champions or runner up. This suggests that a limited number of coaches are likely to get any
value out of these contract clauses, but for those that do, they may be qualifying for two or three
of these payoffs. In other words, excellent on-field performance may result in substantially
increased compensation payments for a small set of coaches.
33
3. Football Coach Academic Performance Incentive Pay
Football coaches also are rewarded for their team’s academic performance off the field.
The academic performance incentives may be present to balance on field incentives. It is hard to
assess how difficult it is for coaches to achieve these goals and whether the coach’s efforts really
have any effect at all on players’ scholastic outputs. While a cynic might claim that these
provisions are included solely to make it appear that universities care about whether their
football players are getting a good education, including them in football coach employment
contracts does create some financial incentives for the coaches to insure that team members
perform well in school.
Table 2D provides information on three different types of academic performance
measures that frequently appear in football coach contracts: APR, Graduation Rate; and GPA. The Academic Progress Rate (APR) is a multi-year measurement of academic progress and an institution’s retention of student-athletes.56 Graduation rate measures the percentage of entering football players that graduate within six years of entering the institution.57 GPA includes any threshold measure of a team’s players’ grade point average. Institutions often create varying incentive payments that increase as performance targets are met. Table 2D summarizes information on these pay for academic performance clauses for the year 2013. It shows the maximum amount58 of academic performance pay as a percentage of each coach’s base salary and, in parentheses, the percentage of employment contracts that contain this particular clause. It is clear from the data that there is substantial variation across conferences in how commonly such provisions appear in coaches’ contracts. While the Pac-12
56 Charlie Zegers, Academic Progress Rate (APR), About.com NBA, available at: http://basketball.about.com/od/collegebasketballglossary/g/Academic-‐Progress-‐Rate-‐Apr.htm 57 This is the standard definition we found for this term. FAR REPORT TO THE CLEMSON UNIVERSITY ATHLETIC COUNCIL, UNDERSTANDING GRADUATION RATES, GRADUATION SUCCESS RATES AND ACADEMIC PROGRESS RATES (April 2007), https://www.clemson.edu/administration/councils/athleticcouncil/documents/understanding_fgr_gsr_apr.doc.
58 We calculated the maximum payoff that could result from the satisfaction of any of these clauses as there is enormous variation in the contracts over the amounts and exact thresholds for payment.
34
conference has the highest use of academic pay for performance clauses for APR and graduation rate, the Big East has the highest usage of GPA provisions.
Table 2D: Academic Performance Maximum Incentive Pay by Conference As a Percentage of
Base Salary (Percentage of Contracts With Clause)
APR Graduation Rate GPA Pac-12 9.6% (39.3%) 15.3% (57.4%) 9% (34.4%) ACC 14.3% (13.4%) 19.7% (44.8%) 0% (0%) Conference USA 10% (32.4%) 17.1% (36.8%) 5.2% (19.1%) Big 12 6.7% (6.6%) 14.4% (29.7%) 11% (2.2%) Big East 4.5% (35%) 11.2% (27.5%) 6.5% (47.5%) Mid-American 4% (26.6%) 3% (21.8%) 2.6% (17.7%) Sun Belt 2.2% (37.3%) 5.5% (18.6%) 2.5% (20.3%) Big Ten 13.7% (9.6%) 10.4% (18.1%) 23.3% (7.5%) Mountain West 3.3% (21.5%) 9.5% (13.9%) 13.6% (4.6%) SEC 23% (23.1%) 16.2% (11.5%) 21.8% (7.7%) Not Yet FBS 2.3% (10.3%) 2.1% (6.9%) 0.5% (3.5%) WAC 10.8% (15.2%) 17.8% (6.3%) 8% (8.9%) Independent 5.7% (3.5%) 0% (0%) 0% (0%)
Looking first at the value of these clauses’ payoffs, we see that a substantial majority of
them pay a maximum amount equal to less than 14% of base salary which is the minimum
possible payment for on-field coach performance shown in Table 2C for any category of
performance in 2013. Similarly if we compare academic clauses with on-field performance
clauses, we see that on-field performance clauses are much more common.59 We conclude that
academic performance clauses are far less economically significant than on-field performance
pay for coaches that are highly successful on the field.60 However, we must be mindful that
many coaches will not qualify for any of the on-field performance payments listed in Table 2C,
59 Some contracts have multiple academic performance hurdles in them so that coaches could receive more than one payment for their players’ academic performance. 60 GRANT ET AL., supra note 10, at 231 (…”academic-‐based payments [to coaches] are trivial compared to bonuses based on winning.”).
35
so that academic performance payments may in practice be more significant than it appears at
first blush, especially to coaches with poor on-field performance records.
4. CEOs’ Compensation Data
We next calculate the executive compensation levels for CEOs at our matched set of
public firms. The basic components of public company executive pay in the U.S. are: salary;
bonus; stock awards (usually restricted stock); stock options; non-equity incentive pay; and a
residual category of other compensation.61 We obtain these values from the Compustat electronic
database. Table 3 reports the 2013 averages for the matching set of CEOs at the comparable
firms, where we have divided the sample up into four quartiles from smallest firms to largest
firms.
Table 3: CEO Compensation 2013 (Dollar Mean Values)
First Quartile Second Quartile Third Quartile Fourth Quartile Salary 402,002 446,875 449,851 511,876 Bonus 88,322 86,660 89,599 128,862 Restricted Stock Awards
176,953 405,627 393,610 823,703
Option Awards 189,078 230,272 274,918 444,623 Non-equity Incentive Pay
147,550 146,508 270,712 261,714
Other Comp 80,723 64,850 79,611 97,004 Total Comp 1,084,630 1,380,796 1,608,303 2,267,785
61 Randall S. Thomas, Explaining the International Pay Gap: Board Capture or Market Driven?, 57 VAND. L. REV. 1171, 1182-‐1186 (2004).
36
The values reported in Table 3 show that average total compensation increases as firm size
increases. Fourth quartile average total CEO pay is $2,267,785, whereas first quartile total
average compensation comes out at $1,084,630, or slightly less than half of that of the largest
firms. In addition, each category of compensation generally increases incrementally as we move
from the smaller firms to the larger firms, although the firms in the third quartile show some
variation (particularly with respect to restricted stock awards) in this respect. Incentive based pay
becomes more important at the larger firms, particularly stock options and restricted stock
awards. There is relatively little change in salary levels as firm size increases.
5. Comparing CEO and Football Coach Compensation Packages
How do the overall size and growth rates of the different components of CEO and college
football coach compensation compare? The first significant difference is the relatively large
fixed pay component to college football coaches’ compensation compared to CEO pay. In
contract year 2013, Table 2A shows that football coaches are paid over $1.5 million in fixed pay
with base salary constituting roughly 35% of that total. While we cannot calculate total
performance based pay using solely the information in the coach employment contracts, we can
see from Table 2C that the most a coach could earn in any one category would be for winning
the national championship and that would result in their getting 94% of their base salary. The
maximum that they could earn for satisfying any one of the academic performance off-field
bonuses would be another approximately 22% of their base salary. If we total those bonuses
together, we see that performance based pay for football coaches is significantly less than fixed
pay with a maximum of perhaps 50% of fixed pay.
Table 3 shows that for a comparable sized set of public companies, CEOs earn most of
their income (roughly 75%) from performance based pay and only about a half million in fixed
pay (about 25% of their total pay). In a very crude comparison, this is roughly the reverse of
what we see for college football coaches.
37
One potential explanation is that coaches have much shorter job tenure and more frequent
contract re-negotiations. A successful coach will promptly receive an increase in pay and a new
contract. This makes it much less important to include performance based pay in the original
contract. Conversely, poor performance is likely to lead to prompt termination. Compensation
for termination will be included in the severance provisions of the initial contract, making it less
important to worry about the downside of performance based pay as well.
A second important point is that when we compare Table 2B and Table 3 we see that the
compensation quartile differentials are very close to the same for both football coaches and
CEOs, except for the smallest schools. In other words, the firm size effects in executive
compensation are very close to the program size effects for college football coaches in the top
three quartiles of the compensation range for a matched set of firms and universities. This
suggests that the two types of position have some close similarities.
Another important comparison is the rate of growth of pay for the two groups. Our data
in Table 2A show that from contract year 2005 to contract year 2013, fixed pay more than
doubled. Earlier data confirm this trend: in 1999, five coaches earned more than $1 million,
while in 2008 the average salary for NCAA Division I Football Bowl Subdivision (FBS) coaches
was $950,000 per year with over 40 coaches earning more than $1 million.62 Given the relatively
constant percentage of base salary represented by the performance pay shown in Tables 2C and
2D, it seems reasonable to conclude that total football coach pay has doubled over our sample
period.
Research on CEO pay shows that from 1992 to 2001, CEO median pay increased by
213%.63 This doubling of CEO pay is consistent with what we are observing with college
football coaches. However, in more recent years, college football coach compensation is rising
62 Christian Dennie, There are No Handshake Deals in College Coaching Contracts, 20 No. 1 ANDREWS ENT. INDUS. LITIG. REP. 2, *1 (2008) (“[T]he average head football coach in the six major conferences . . . earns approximately $1.4 million per year.”). 63 Carola Frydman & Dirk Jenter, CEO Compensation, 2 ANN. REV. FIN. ECON. 75, 78 (2010).
38
much faster than CEO pay: from 2007 to 2011 CEO pay increased 23%, while over that same
time period college football coach pay increased 44%.64
There are several factors that explain these differences. First, college football television
revenues are growing rapidly in recent years and our theory predicts that coaches’ capture a
larger percentage of value created. Second, college football coaches have shorter job tenure than
CEOs and this is likely to put pressure on them to get paid as much as possible early in their
contracts. Third, successful football coaches have very frequent contract amendments as their
home schools try to persuade them to remain in place, while other schools try to lure them away.
A final reason for the observed differences in compensation may stem from the fact that coaches’
success (and failure) is more readily observed than with CEOs, making it more apparent who
should be paid more. All of these drivers for coach compensation increases put upward pressure
on compensation levels.
C. Contract Length and Duties
Contract length is another common element in the employment agreement in both coach and
CEO contracts. Table 4 compares the distribution of contract lengths from our sample of coaches
with CEOs.65 Earlier work found that while the bulk of CEO employment contracts are for a
definite term, many CEO contracts do not specify a fixed term of years in them.66 The third
column of Table 4 presents data showing the percentage of CEO contracts (totaling 87%) that
did contain a definite term. The remaining contracts (13%) did not contain a definite term.67 By
comparison, virtually all football coach contracts have a fixed term of years.
64 Brady et al., supra note 9. 65 For purposes of comparing these data with the contract lengths for CEO contracts, we use the results of prior research by one of the authors. Schwab &Thomas, supra note 6, at 247. 66 Id. at 247. 67 This category includes contracts where length was not mentioned (2.9%), contracts lasting until the CEO was terminated (8.8%); and contracts lasting until the CEO is 60-‐65 (1.3%).
39
Table 4: Contract Length Comparisons (Percentages)
Contract Length in Years
Coaches CEOs
Indefinite Term or Term Not Mentioned
0.75% 13%
1 3.74% 5.1% 2 2.46% 9.6% 3 8.65% 36.0% 4 18.38% 6.4% 5 41.67% 24.0% >5 24.36% 5.9% Percentage of Contracts with a Fixed Term
99.25% 87%
Football coach contracts are typically longer than CEO contracts. More than 65% of
football coaches have contracts that are 5 or greater, while approximately 30% of CEOs have
contracts that long. The most frequent CEO contract length is three years. The principal
explanation for this difference appears to be that football coaches must be able to credibly
commit to potential recruits that they will be coaching at the school for the entire period of time
that the recruit will attend it.68 Thus, a high school junior or senior wants to believe that the
coach who recruits them is contractually committed to remain there for at least the next four
years.
CEO contracts, on the other hand, reflect market forces at work in that labor market. In
particular, the internal labor market for managers is often compared to a tournament, where the
prize for winning is to be promoted to the next level of management. In order for the tournament
to be effective, there must be openings created at the higher levels of management within a 68 Sammi Bjelland, Pay Day for College Coaches, KELOLAND.COM, May 11, 2014, http://www.keloland.com/newsdetail.cfm/pay-‐day-‐for-‐college-‐coaches/?id=164254 (other schools will use a coaches’ short term contract as a scare tactic to convince recruits that the coach won’t be there for long); Jeff Barker, Friedgen’s firing was weeks in the making, BALT. SUN, December 21, 2010, http://articles.baltimoresun.com/2010-‐12-‐21/sports/bs-‐sp-‐terps-‐fridge-‐1221-‐20101220_1_lame-‐duck-‐atlantic-‐coast-‐conference-‐coach-‐football-‐coach (lame duck coaches cannot recruit effectively).
40
reasonable period of time to incentivize lower level managers to remain with the corporation and
continue to compete. Perhaps as a result, CEO employment contracts appear to be shorter than
football coach contracts.
Another explanation offered in the literature focuses on the difference between inside and
outside hires. This research finds that CEOs hired from outside of the firm have on average one
year longer contracts than CEOs hired from internal positions.69 This is true because outside
CEOs face greater uncertainty about their long term future with the company and therefore
negotiate for greater contractual protection.70 If this is correct, then we would anticipate that, on
average, football coaches, who are almost always hired from outside, would have longer
contracts than CEOs who are frequently internal hires.
When it comes to the articulation of duties, the difference between the two types of contracts
is quite stark. Coach’s contracts typically contain a much more detailed set of job duties than
those in a CEO contract. For example, a typical CEO contract will have a short, one paragraph
(or less) general description of their responsibilities that is vague and unspecified.71 By
comparison, the employment contracts of college football coaches typically give a very detailed
and extensive definition of the duties expected of the coach.72 For instance, Alabama Coach
Nick Saban’s recent employment contract contains a four page list of his job duties and
responsibilities.73 In part, this reflects the expanding role that coaches are expected to play in
69 Gillan, et al., supra note 9, at 1631. 70 Id. 71 For example, in CEO Raymond Gilmartin’s employment contract with Merck & Co., Inc. (the drug company), there is a single sentence that describes his job duties: “Employee shall have the powers, responsibilities and authorities of President, Chief Executive Officer and Chairman of the Board of Directors of the Company as established by custom and practice.” Gilmartin Employment Agreement with Merck & Co., Inc., dated June 9, 1994. 72 Greenberg, supra note 13, at 147. This definition of duties includes things such as: Coaching responsibilities; Recruiting responsibilities; TV/Radio requirements; Shoe/Apparel sponsorship requirements; Public Appearance requirements; Compliance with NCAA Rules and Regulations; Compliance with University Rules and Regulations; and the oversight of Coaching Staff/Student Athletes. 73 Amended and Restated Head Coach Employment Contract, Between The Board of Trustees of the University of Alabama and Nick L. Saban, section 2.02 Duties and Responsibilities, effective February 1, 2013.
41
recent years: instructor, “fund-raiser, recruiter, academic coordinator, public figure, budget
director, television, radio and internet personality, and alumni glad hander.”74
D. Termination Clauses: With Cause and Without Cause
Having a long term contract, however, does not mean that a CEO or a football coach has any
guarantee of being employed for the full length of the contract. In fact, any employee can be
terminated by their employer at any time.75 The question becomes what does the employee
receive from the employer if they are fired? Employees that are fired with cause, that is for
committing an act or omission that is specified in their agreement as grounds for cause
termination, are typically given little or no compensation after their termination. By comparison,
an employee that is fired without cause -- for any reason not specified as grounds for termination
for cause in their employment contract -- is generally provided with some type of severance
package. In this section, we explore the definitions for cause in the CEO and football coach
contracts and the amount of severance payments that each one is entitled to when terminated
without cause.
1. For Cause Terminations
From the firm or university’s perspective, a for cause termination clause is necessary to
protect its image and reputation from damage caused by an employee’s misconduct.76 However,
for cause terminations normally create immediate financial hardship and damages for employees.
As a result, coaches and executives are concerned that they not be unjustifiably, or subjectively,
74 Greenberg, supra note 13, at 130. 75 Greenberg, supra note 13, at 245 (“The university, therefore, may remove the coach from his position at any time, with or without a valid reason.”). 76 Adam Epstein, An Exploration of Interesting Clauses in Sports, 21 J. LEGAL ASPECTS SPORT 5, 17 n. 65 (2011).
42
fired for cause. In theory, this should make negotiations over the items that are included in the
employment contract’s definition of “cause” contentious. However, for CEOs, this is not the
case.77 In fact, the items enumerated as constituting cause are relatively limited in CEO
contracts and are often taken directly from the prior CEO’s contract. In practice, CEO’s are
almost never fired for cause, with the exceptions arising out of situations involving flagrant
abuses, such as blatant sexual harassment cases.
Things are more complicated for football coaches. “Coaching is a job where you are
hired to be fired. The back end of the deal is as important as the front end of the deal. The first
day of the job must be spent planning for the last day of the job.”78 Other researchers have found
that approximately 1 in 10 teams annually change head football coaches for performance
reasons.79 With the high stakes of college football, coaches are often “fired or pressured to resign
as a way of ‘fixing’ [their] programs.”80 With termination likely from the start of the
relationship, it becomes crucial to the coach to distinguish what constitutes cause from what does
not.
Furthermore, as the media coverage of college football has expanded so too has media
coverage of coaches. Coaches face a growing risk of termination for purely personal conduct, not
involving on field performance issues.81 As a result, it is common for coaching contracts to
contain termination provisions pertaining to moral turpitude, perpetuation of willful fraud,
conduct seriously prejudicial to the best interests of the university, immoral acts, habitual
77 Schwab & Thomas, supra note 6, at 238. 78 Martin J. Greenberg, Representation of College Coaches in Contract Negotiations, 3 MARQ. SPORTS L.J. 101, 103 (1992). 79 E. Scott Adler et al., Pushing “Reset”: The Conditional Effects of Coaching Replacements on College Football Performance, 94 SOC. SCI. Q. 1 (2012).
80 Id. at 1–2. Whether poor performance constitutes grounds for termination with cause depends crucially on the language of the coach’s employment contract. For further discussion, see Table 5A below.
81 Dennie, supra note 50, at *2. For example, the University of Arkansas recently publicly dismissed its head football coach Bobby Petrino for purely personal reasons. Petrino was very successful on the field and was dismissed on the heels of a 2012 Cotton Bowl win (11-‐2 season record) and at the time of his dismissal the team was considered a national title contender. See Joe Schad, Bobby Petrino emotional, regretful, ESPN, Aug. 10 2012, http://espn.go.com/college-‐football/story/_/id/8251166/bobby-‐petrino-‐fired-‐arkansas-‐razorbacks-‐coach-‐apologizes-‐interview-‐espn.
43
intoxication, dishonesty and gross negligence in performance.82 Coaching contracts also
frequently contain for cause termination provisions related to program violations of NCAA rules
and regulations.
To be terminated for cause requires complimentary “due process clauses” providing the
specific requirements of notice of any hearings afforded to the coach at termination or
suspension.83 Generally, due process clauses are drafted to provide coaches with a specter of
rights along a continuum. Coaches receive the same due-process rights as other university
employees or they may have the opportunity to appeal to the athletic director or an independent
arbitrator.84
Table 5A provides a breakdown of the items listed as independent grounds for cause
termination in college football coach contracts. There is substantial variation amongst the
contracts as to the grounds for termination with cause. But all of these contracts contained
detailed provisions governing for cause terminations.85
82 Dennie, supra note 50, at *3; Greenberg, supra note 13, at 211-‐213. 83 Dennie, supra note 50, at *4. “A well-‐drafted due-‐process clause includes written notice to the coach of contemplated suspension or termination; the opportunity for the coach to defend himself before an impartial arbitrator; clear procedures for notice, rules of evidence, the right to cross-‐examine witnesses, and the right to an attorney during the hearing and discovery phase; the effect of a finding of just cause; and appellate rights.” Id. 84 See id.; see also Brent C. Moberg, Navigating the Public Relations Minefield: Mutual Protection Through Mandatory Arbitration Clauses in College Coaching Contracts, 16 J. LEGAL ASPECTS SPORT 86 (2006) (regarding mandatory arbitration clauses). 85 For the college football coach employment contracts, we used the definitions set forth below. Gross Misconduct was coded as such if it was listed as grounds for termination with cause. We coded as Neglect provisions that stated that neglect of duties, failure to perform duties, inability to perform duties, or absence from duties, constituted grounds for cause termination. Breach of Contract provisions providing that a breach of the employment contract was grounds for termination with cause were coded as Breach of Contract. Whenever the conviction of a crime constituting a felony was grounds for termination for cause, we coded that as Felony. Similarly, conviction of a crime involving moral turpitude, or any other provision relating to moral turpitude – including things such as violation of anti-‐discrimination policies or sexual harassment, were coded as Moral Turpitude. Fraud or dishonesty was listed as grounds for termination with cause in some contracts. Unprofessional Conduct is defined as unprofessional behavior or other conduct that reflects negatively upon the university. Violation is a violation of NCAA, Conference, or University rules; the failure to report a violation of NCAA, Conference, or University rules; the counseling of those under the coach’s supervision to fail to respond promptly and accurately to any NCAA, Conference, University, or other official inquiry; or the failure by the coach to respond promptly and accurately to any NCAA, Conference, University, or other official inquiry. Gambling or betting of any kind is another ground for termination with cause in some contracts. Substance Abuse of any kind is another ground for termination with cause sometimes. The existence of a conflict of interest can be grounds for cause
44
Table 5A: College Football Coach Contract for Cause Provisions (Percentages of Contracts)
Provision Percentage of Contracts With Provision
Gross Misconduct 85% Unprofessional Conduct 65.4% Neglect of Duties 82.3% Breach of Contract 60.8% Felony 31.1% Moral Turpitude 50.5% Fraud or Dishonesty 30.5% Violation of NCAA rules 93.1% Gambling 11.7% Substance Abuse 15.1% Conflict of Interest 4.6% Drug Test Program Failure 7.7%
The CEO contracts also generally contained provisions on termination for cause, but they
tended to be less numerous and less detailed. Table 5B contains a list of the terms used in the
CEO contracts and the percentages of contracts in which they appeared.
Table 5B: CEO Contract for Cause Provisions (Percentages of Contracts with Provision)
Provision Percentage of Contracts With Provision
Not mentioned 2.4% Mentioned but not specified 4.5% Fiduciary breach 50.7%
termination. Finally, the failure to cooperate in the enforcement and implementation of any drug testing program can in some instances be grounds for cause termination.
45
Willful misconduct 69.1% Gross misconduct 39.2% Incompetence 3.5%
Moral turpitude 72.2% Failure to Perform Duties 57.8% Substance Abuse 4.8% Sexual Harassment 0.5%
A close comparison of Tables 5A and 5B reveal many similarities, but also some important
differences. In terms of similarities, both types of contracts typically define gross/ willful
misconduct,86 or commission of an act of moral turpitude (including sexual harassment), as
grounds for termination for cause, although coaches’ contracts seem more lenient on moral
turpitude violations. There seems little doubt that these actions should constitute sufficient
grounds for penalizing the employee by terminating them for cause.
Less frequent, but still common, is language defining cause as a breach of the employment
contract (coaches) or a breach of fiduciary duty (CEOs). These are not the same thing and it is
interesting to think about why they differ. Since the CEO is a fiduciary to his or her corporate
employer, whereas the coaches are not, it would make little sense to include breach of fiduciary
duties as grounds for cause termination in coaches’ contracts. However, CEO’s are contracting
parties with their firms, just as coaches are, which raises the question why CEOs’ contracts do
not include breach of contract as grounds for cause termination? One possible explanation is
path dependence: prior CEO contracts did not contain such language, and the parties just include
the provisions from prior contracts in new agreements.87 Another reason could be the
differences in the two types of contracts in their definitions of employee duties: coaches have
very specific, detailed lists of duties whereas CEOs do not, making it easier for a coach to violate
their contracts by failing to perform specified duties. Nevertheless, the difference is puzzling.
86 These would presumably include the commission of fraud or a felony by the coach or CEO. 87 Path dependence has been found to be a powerful explanatory variable for the use of arbitration provisions in CEO employment contracts in earlier work by one of the authors. Randall Thomas et al., Arbitration Clauses in CEO Employment Contracts: An Empirical and Theoretical Analysis, 63 VAND. L. REV. 959, 997-‐998 (2010).
46
A second intriguing difference is the relative frequency of contract provisions defining
neglect of duties as grounds for termination for cause. Football coaches’ contracts almost
universally contain such provisions (82.3%), whereas CEO contracts include them less often
(61.3% if we combine the categories incompetence and failure to perform duties). This suggests
that in a significant number of instances it will be easier to terminate a college football coach for
poor job performance, although such clauses have not generally been interpreted as permitting
for cause termination for a poor win loss record.88
The remaining differences arise from particular characteristics of college football coaches’
jobs. For example, all football coach contracts define cause to include violations of NCAA
rules.89 Obviously, CEOs do not have to worry about such compliance. However, gambling and
substance abuse provisions seem to be much more prevalent in football coach contracts than in
CEO contracts. In part, this may be because of the importance that universities place on football
coaches being models of good behavior for student athletes.90 A second aspect of these
provisions may relate to concerns that NCAA rules prohibit gambling and schools want to
remove any ambiguity about that point in these contracts.91
2. Without Cause Terminations
In contrast to cause terminations where the departing employee receives little or nothing
from the employer, terminations without cause for CEOs and football coaches result in
88 Greenberg, supra note 13, at 226. 89 This category includes a violation of NCAA, Conference, or University rules; the failure to report a violation of NCAA, Conference, or University rules; the counseling of those under the coach’s supervision to fail to respond promptly and accurately to any NCAA, Conference, University, or other official inquiry; or the failure by the coach to respond promptly and accurately to any NCAA, Conference, University, or other official inquiry. 90 Kai Sato, In Defense Of College Football Coaches, HUFF POST SPORTS, January 10, 2014, http://www.huffingtonpost.com/kai-‐sato/in-‐defense-‐college-‐football-‐coaches_b_4570443.html (“Good college football coaches are often role models, mentors and father figures for their players.). 91 Martin J. Greenberg, Termination of College Coaching Contracts: When Does Adequate Cause to Terminate Exist and Who Determines Its Existence?, 17 MARQ. SPORTS L. REV. 197, 231-‐233 (2006).
47
substantial payouts to the departing personnel. In the case of football coaches, they typically
receive an amount equal to net present value of the compensation that they would have received
if they had remained as a coach until the end of their contract.92 For example, if a coach had a
five year contract, and was terminated after their first year as a head coach, then they would be
entitled to receive four years of compensation payments.93
Employment contract provisions related to without cause severance payments for CEOs
are significantly less generous and typically only pay them a multiple of base salary. Table 6
provides some detail on the actual distribution of the amounts of these payments listed in the
employment contracts for both football coaches and CEOs.94
Table 6: Payments When Terminated Without Cause (Percentages of Base Salary)
Payments Coaches (Total Compensation)
CEOs (base salary)
Payment = 0 12.5% 1.9% 0< Payment < 1 Year 1.6% 2.1% Payment = 1 Year 2.0% 8.0% 1< Payment < 2 Year 5.2% 2.4% Payment = 2 Year 2.3% 29.1% 2< Payment< 3 Year 5.9% 2.9% Payment = 3 Year 3.3% 20.5% 3< Payment < 4 Year 8.8% 2.7% Payment = 4 Year 5.9% ----- 4< Payment < 5 Year 11.9% ----- Payment = 5 Year 11.5% ----- Payment > 5 Year 29.0% ----- 92 Dennie, supra note 50, at *6. 93 In some cases, however, these payments are made on the same time schedule as provided in the contract with the departed head coach subject to a duty to mitigate his damages. For instance, Darrell Hazell’s September 27, 2013 employment contract with Purdue contained such a clause. (§5.2.1.1 Mitigation/Offset stating that: “in the event Purdue terminates Coach’s employment without Cause, Coach shall use reasonable efforts to mitigate Purdue’s obligations hereunder by obtaining football coaching employment…) 94 Ex post agreements to pay additional amounts of compensation after termination are also sometimes entered into for a variety of reasons. See David Yermack, Golden Handshakes: Separation Pay for Retired and Dismissed CEOs, 41 J. ACCT. & ECON. 211 (2006).
48
Not Mentioned or Other Form
----- 32.3%
One likely explanation for the larger payments that are made to football coaches is the
greater likelihood that they will be terminated by their universities and the need to compensate
them for any reputational and financial harm that may result from it. As noted earlier, college
football coaches are more likely to be dismissed than CEOs.
Executives can also choose to leave their job for another employer. Voluntary separation
can either be for good reason or without good reason. CEO voluntary terminations without good
reason usually result in the loss of severance benefits for the executive, so they want to have as
broad as possible a definition of good reason as possible.95 Companies have the opposite
perspective, and want to limit good reason to a narrow set of circumstances to avoid having their
CEO leave easily for alternative employment. This can lead to hard bargaining over, for
example, what constitutes a diminution of duties or responsibilities that could trigger a good-
reason departure. The employment contracts of CEOs do not normally contain a buyout
provision if the CEO chooses to resign without good reason.
In sharp contrast, the employment contracts of college football coaches may contain a
buyout provision if the coach chooses to resign without good reason.96 The payment of this
buyout is frequently conditioned upon the coach resigning to accept another coaching position.
For example, Richard Rodriguez’s August 21, 2007 amended employment contract with West
Virginia University contained a buyout clause that required him to pay the university $4 million
if he left the university prematurely to go to another team. After Rodriguez left to go to
Michigan, he paid $1.5 million and Michigan paid the remaining $2.5 million of the buyout to
West Virginia.97 These provisions provide the university with some compensation for the harm
that they suffer when a coach prematurely departs for another school.
95 Good reason for quitting is commonly defined to include the CEO getting paid less, being transferred to a distant location, or being forced to take on less job responsibility. Schwab & Thomas, supra note 6, at 253. 96 Where no buyout clause is included in the contract, universities typically allow the coach to depart without seeking damages for the harm caused by his departure. Greenberg, supra note 13, at 248. 97 Associated Press, Michigan to pay $2.5M, Rodriguez $1.5M to satisfy WVU buyout, ESPN, July 9, 2008, http://sports.espn.go.com/ncf/news/story?id=3479493.
49
E. Non-Competition Clauses and Arbitration Provisions
Two other important types of clauses that appear frequently in CEO employment contracts
are non-competition clauses and arbitration provisions. Prior research on CEO employment
contracts has found that arbitration provisions appear more than 50% of the time,98 whereas non-
competition clauses are present in around 80% of these agreements.99 Such contract provisions
are uncommon in college football coach contracts.
1. Non-Competition Clauses
Non-competition provisions, and other post-employment restrictions, are very common in
the employment contracts of CEOs. At public companies, non-compete clauses are hotly
negotiated. Companies are very sensitive to the potential for CEOs to walk away from the firm
with proprietary knowledge and go to work for competitors. Non-compete clauses are costly to
the executives where they may be forced into long periods of unemployment in a field where
they have been most successful. These competing interests must be reconciled with the end
result often reflecting the company’s willingness to offer extended periods of severance
payments and judicial decisions limiting the enforceability of strong non-compete agreements.
By comparison, non-competition provisions are exceedingly rare in the employment
contracts of college football coaches. In fact, we found such clauses in just 23 of the contracts of
the 429 coaches in our sample. Coaches are strongly opposed to having such clauses in their
contracts because if they are written broadly, they might not be able to work for another school.
A second reason for their scarcity is that college football coaches’ skills are less firm specific and
98 Erin O’Hara O’Connor, et al., Customizing Employment Arbitration, 98 IOWA L. REV. 133, 161 (2012). 99 Randall S. Thomas et al., An Empirical Analysis of Non-‐Competition Clauses and Other Restrictive Post-‐Employment Covenants, VAND. L. REV. (forthcoming 2015)(manuscript at 35, Table 2, on file with authors).
50
involve less confidential information than those of CEOs. In other words, the university has less
concern that a coach will be able to leverage information in another university and be able to
compete with the school on an unfair basis. The one exception is where the coach has been
actively recruiting a particular set of players to come to the school, leaves and tries to take the
players with him to his new school.100 Here a variation on non-competition clauses, the so-called
“Non-Recruit” clause, stops a departing coach from recruiting any players that he had recruited
for his prior employer.101
A third reason for the relative scarcity of these clauses in football coach contracts may be
that the bargaining power of successful coaches is keeping them from being widely adopted.102
For example, when such a clause appears in a coach’s initial contract with their university, the
coach will bargain hard to get it removed as quickly as possible.103
2. Arbitration
“[A]rbitration is a creature of contract” and may be contained in an employment contract to
avoid the difficulties of publicly resolving scandal and/or conflict.104 CEO employment
100 For example, when Vanderbilt head football coach James Franklin left Vanderbilt to go to Penn State, he took a large number of players that he had recruited for Vanderbilt with him to Penn State. Jeff Lockridge, Analyst Says James Franklin could destroy Vanderbilt recruiting class, USA TODAY SPORTS, January 13, 2014, http://www.usatoday.com/story/sports/ncaaf/2014/01/13/analyst-‐says-‐james-‐franklin-‐could-‐destroy-‐vanderbilt-‐recruiting-‐class/4457185/ 101 Gina A. Pauline & John T. Wolohan, An Examination of the Non-‐Recruit Clause in Intercollegiate Coaching Contracts, 21 J. LEGAL ASPECTS SPORT 219, 228 (2012) (providing review of the enforceability of non-‐recruit clauses in the context of Marist College v. Brady, 924 N.Y.S.2d 529 (2011)). For example, Ohio State University inserted this type of clause in the contracts of its head coaches Jim Tressel, Luke Fickell and Urban Meyer. 102 Michael P. Elkon, Enjoining Nick Saban: Non-‐Compete Agreements and College Football Coaches, NONCOMPETENEWS.COM, Jan. 9, 2012, http://www.noncompetenews.com/post/2012/01/09/Enjoining-‐Nick-‐Saban-‐Non-‐Compete-‐Agreements-‐and-‐College-‐Football-‐Coaches.aspx . 103 For example, Pat Hill’s employment contract dated January 1, 2011 with California State University at Fresno eliminated a non-‐competition clause that had been present in his initial contract dated September 1, 1999. 104 Brent C. Moberg, Navigating the Public Relations Minefield: Mutual Protection Through Mandatory Arbitration Clauses in College Coaching Contracts, 16 J. LEGAL ASPECTS SPORT 86, 106 (2006) (quoting STEVEN C. BENNETT, ARBITRATION: ESSENTIAL CONCEPTS 59 (2002)). Moberg’s article identifies issues in drafting of mandatory ADR clauses in coaching contracts and provides a mock draft of one such clause and confidentiality agreement. Id. at 106–12.
51
contracts frequently contain an arbitration provision.105 Arbitration clauses can be viewed as
desirable by both parties to a CEO employment contract. Companies universally want to
arbitrate disputes to keep things private, and thereby avoid adverse publicity over a messy
termination and possible public litigation. Normally, employees would want to preserve their
right to a jury trial, calculating that a jury of their peers would be more sympathetic to their
situation than to the company firing them. However, CEOs may have good reason to believe that
juries will not identify with their compensation demands, as the amounts involved may seem
excessive to most members of the public. This may lead executives to favor arbitration
generally, although they may still carefully negotiate the selection process for the arbitrators and
their right to appeal from an adverse decision.
The employment contracts of college football coaches infrequently contain an arbitration
provision.106 For our sample, we find arbitration provisions in roughly 11.4% of all of these
contracts. They are sometimes included because coaches and universities are forced to balance
the divergent interests of different stakeholders in an athletic program, so “it is important to
consider methods of dispute resolution that can provide some level of control over a situation
before it becomes a scandal.”107 Placing mandatory arbitration provisions in coaching contracts
has been proffered as a way to keep disputes outside the “spotlight of a courtroom,” as quarrels
between coach and university often draw a significant amount of negative attention to parties.108
However, it appears that they are substantially less popular in coaches’ contracts than for CEOs.
We speculate that path dependence may be a factor in slowing the adoption of these provisions
for football coaches.
IV. Conclusions
105 Thomas et al., supra note 73, at 979. 106 Moberg, supra note 91, at 86. 107 Id. at 86. 108 Moberg, supra note 91, at 86 & 97–98. The authors suggest to further advance the goal of discretion a mandatory ADR clause might also be paired with a confidentiality provision. Id. at 88.
52
We conclude that college football coaches’ employment contracts are quite similar to
CEO contracts in that while the dollar value of their compensation is high, as long as it is
structurally aligned with value creation for shareholders, the level may be warranted. Moreover,
we find, as theory would predict, that college football coaches have shorter job tenure, contracts
that focus more on termination events and payoffs, and contracts that contain bigger severance
payments for departing coaches.
We also predicted that coaches would receive more performance based pay and have
shorter employment contracts than CEOs. We find that successful coaches are commonly
rewarded with an amended contract and that as a result their contracts are effectively in place for
a shorter time period. While coaches are less likely to have large pay for performance payments
in their initial contracts, successful coaches promptly receive pay increases and a new contract,
while unsuccessful coaches are terminated. Economic theory also led us to predict that football
coaches will see their compensation rise more rapidly and that coaches will capture a greater
percentage of increases in program value, especially for external hires. We found that
compensation of coaches has been rising faster than for CEOs recently, but can offer only
circumstantial evidence that they are capturing more of the increase in program value than CEOs
are capturing in firm value.
At the same time, there are several differences between CEO and football coach contracts
that require further analysis to explain. First, only 50% of CEOs have employment contracts,
while virtually all football coaches have them. While many CEOs have some other form of
contractual arrangements with their firm, these agreements are not as protective of these
employees as a written employment contract. One possible explanation is that coaches’ success
(and failure) is more readily observed than with CEOs. Both universities and coaches have
strong interests in careful planning for the coaches’ more frequent job changes. This may be
coupled with universities need to have well-specified rights to terminate coaches for NCAA rule
violations, as discussed below in section. Another possible explanation is that internal
candidates are less likely to request written contracts because of their greater knowledge of the
53
company’s board of directors and CEOs are more likely than football coaches to be drawn from
the internal ranks.
Next, the vast majority of CEO contracts have non-competition clauses in them, whereas
almost no football coach contracts have them. This difference may reflect that coaches’ skills
are generally transferable to any football program and there is little loss to the university of firm
specific capital when a coach leaves. However, a non-compete clause effectively forecloses a
coach’s labor market opportunities so they will resist it strenuously. The one exception where the
departing coach may be able to unfairly compete with his former employer is with respect to
incoming recruits. Based on their pre-existing personal relationships, coaches may have an
advantage in convincing new recruits to abandon their former commitments to come join them at
their new school. To stop this form of competition, some universities have insisted on “Non-
Recruit” clauses, a form on non-competition provision.
Finally, while arbitration provisions are quite common in CEO employment contracts,
they are included in little more than 10% of football coach contracts. This is puzzling as many
of the same considerations that lead firms to adopt arbitration provisions for CEO agreements
would appear to be present for football coaches. We expect that over time this type of provision
will become more common in football coach contracts, but that path dependence has slowed
down its adoption.
In sum, we find that a close comparison of CEO employment contracts and college
football coach contracts shows that both sets of agreements conform to the predictions of
economic theory. Thus, if one believes that CEO compensation is set by the market at an
appropriate level, and that employment contracts reflect this equilibrium, then one should reach
the same conclusion about football coaches.